Anti-WAR protest

[jojo]

I agree with this. I can understand performance in relation to a mean of all players at that position in that aspect of the game. I have trouble then relating that to wins for the player's team. If a team's roster were filled entirely with 0 WAR players, offensively and defensively, should that team finish .500?

No because 0 WAR is replacement level which is decidedly less than .500.

[BluegrassRedleg]

What am I missing on Blyleven's 1973 season? Doesn't seem overly impressive alongside some of those others.

[dougdirt]

What am I missing on Blyleven's 1973 season? Doesn't seem overly impressive alongside some of those others.

325 innings and a 3.74 K/BB goes a very long way.

[westofyou]

325 innings and a 3.74 K/BB goes a very long way.

it's an interesting benchmark

INNINGS PITCHED >= 300
STRIKEOUTS/WALKS >= 3.70
RSAA displayed only--not a sorting criteria
 ERA			            YEAR	    DIFF	  PLAYER	LEAGUE	  IP	  SO/BB	 RSAA
15	 Addie Joss		     1908	    1.22	    1.16	  2.39	 325	   4.33	    47
12	 Bert Blyleven		     1973	    1.30	    2.52	  3.82	 325	   3.85	    53
4	 Bob Gibson		     1968	    1.86	    1.12	  2.98	 305	   4.32	    56
10	 Christy Mathewson	     1911	    1.40	    1.99	  3.39	 307	   3.71	    47
13	 Christy Mathewson	     1912	    1.28	    2.12	  3.40	 310	   3.94	    45
18	 Christy Mathewson	     1913	    1.14	    2.06	  3.20	 306	   4.43	    40
22	 Christy Mathewson	     1908	    0.92	    1.43	  2.35	 390.2	   6.17	    44
1	 Cy Young		     1901	    2.04	    1.62	  3.66	 371.1	   4.27	    72
24	 Cy Young		     1903	    0.87	    2.08	  2.95	 341.2	   4.76	    44
25	 Cy Young		     1905	    0.82	    1.82	  2.65	 320.2	   7.00	    30
31	 Cy Young		     1904	    0.63	    1.97	  2.60	 380	   6.90	    29
19	 Denny McLain		     1968	    1.02	    1.96	  2.98	 336	   4.44	    42
30	 Don Drysdale		     1963	    0.66	    2.63	  3.29	 315	   4.40	    14
14	 Ed Walsh		     1910	    1.25	    1.27	  2.52	 369.2	   4.23	    42
21	 Ed Walsh		     1908	    0.97	    1.42	  2.39	 464	   4.80	    39
26	 Ferguson Jenkins	     1974	    0.80	    2.82	  3.62	 328.1	   5.00	    23
27	 Ferguson Jenkins	     1971	    0.70	    2.77	  3.47	 325	   7.11	    39
29	 Ferguson Jenkins	     1970	    0.66	    3.39	  4.05	 313	   4.57	    53
33	 Ferguson Jenkins	     1969	    0.39	    3.21	  3.60	 311.1	   3.85	    22
34	 Ferguson Jenkins	     1968	    0.35	    2.63	  2.98	 308	   4.00	    28
6	 Grover C Alexander	     1915	    1.52	    1.22	  2.74	 376	   3.77	    69
16	 Jim Bunning		     1966	    1.20	    2.41	  3.61	 314	   4.58	    43
28	 Jim Kaat		     1966	    0.69	    2.75	  3.44	 304.2	   3.73	    37
8	 Juan Marichal		     1969	    1.50	    2.10	  3.60	 300	   3.80	    51
11	 Juan Marichal		     1966	    1.38	    2.23	  3.61	 307	   6.17	    42
23	 Juan Marichal		     1963	    0.88	    2.41	  3.29	 321	   4.07	    26
32	 Juan Marichal		     1968	    0.56	    2.43	  2.98	 326	   4.74	    16
20	 Rube Waddell		     1904	    0.98	    1.62	  2.60	 383	   3.84	    40
3	 Sandy Koufax		     1966	    1.88	    1.73	  3.61	 323	   4.12	    58
7	 Sandy Koufax		     1965	    1.50	    2.04	  3.54	 336	   5.38	    40
9	 Sandy Koufax		     1963	    1.41	    1.88	  3.29	 311	   5.28	    40
2	 Walter Johnson		     1912	    1.95	    1.39	  3.34	 369	   3.99	    74
5	 Walter Johnson		     1913	    1.78	    1.14	  2.92	 346	   6.39	    75
17	 Walter Johnson		     1910	    1.16	    1.36	  2.52	 370	   4.12	    49

[cincinnati chili]

Honestly, I think we need to get away from trying to shoehorn the entire game into one number that rules them all.

I don't have a problem with trying if the number is reliable. Pitching WAR is more reliable than a lot of stats that get thrown around all the time, but it also has its problems. For one, the impact of quality relievers is undervalued in relation to middle of the road starting pitchers.

[RedsManRick]

I don't have a problem with trying if the number is reliable. Pitching WAR is more reliable than a lot of stats that get thrown around all the time, but it also has its problems. For one, the impact of quality relievers is undervalued in relation to middle of the road starting pitchers.

I would happily disagree with this notion. To the extent that relief pitcher run prevention is more valuable to the team because of timing, I see little reason that the value from that event should be credited to the pitcher whom the team chose to use. Though I recognize many people disagree with that perspective, believing that player performance should be leverage adjusted.

Put differently, if I'm asked which player was more effective in a given year, I'll take the excellent reliever who pitched in long relief over the merely decent one who served as closer, even if leverage adjusted WAR would suggested otherwise. I can decide to use the long reliever in higher leverage situations, making the runs he prevents more valuable. I can't make the lesser pitcher prevent more runs.

Or considered against the mediocre SP, how many more games is your team in a position to win because you have a mediocre SP instead of a poor one? And how does that compare to the effect of relievers of a comparable WAR gap?

To me, the strength of WAR is precisely that it gets us outside of these kind of mental biases we're so prone to.

[757690]

I would happily disagree with this notion. To the extent that relief pitcher run prevention is more valuable to the team because of timing, I see little reason that the value from that event should be credited to the pitcher whom the team chose to use. Though I recognize many people disagree with that perspective, believing that player performance should be leverage adjusted.

Put differently, if I'm asked which player was more effective in a given year, I'll take the excellent reliever who pitched in long relief over the merely decent one who served as closer, even if leverage adjusted WAR would suggested otherwise. I can decide to use the long reliever in higher leverage situations, making the runs he prevents more valuable. I can't make the lesser pitcher prevent more runs.

Or considered against the mediocre SP, how many more games is your team in a position to win because you have a mediocre SP instead of a poor one? And how does that compare to the effect of relievers of a comparable WAR gap?

To me, the strength of WAR is precisely that it gets us outside of these kind of mental biases we're so prone to.

Completely disagree.

The idea of WAR is not to determine who was a better pitcher, who was more effective, but to determine who helped his team win more often. Otherwise, it shouldn't be called Wins Above Replacement, it should called, Run Prevention/Creation Above Replacement. They translate the runs created/prevented into Wins for a reason.

Pitching in high leverage innings means that you are having a greater effect on your team's ability to win. A pitcher who pitches in high leverage situations should get more credit for helping his team win than a pitcher who pitches in low leverage situations. The only reason why we don't need to do that for hitters, is the general assumption is that most hitters hit in the same amount of high and low leverage situations over the course of a season. That is not the case with relief pitchers.

[757690]

I think when one needs such an extreme example in order to make their point, the point they're making is on rather shaky ground.

The funny thing is that this is a terrible example to make his point.

Guidry was 25-3, or had 22 wins overall.

Assuming that the game is 50-50 in terms of run creation and run prevention being responsible for team wins, that would put Guidry at being responsible for 11 wins. Then you have to factor in the value of the defense behind him. A very rough estimate would put it in the 75-25 range between the pitcher and his defense being responsible for run prevention. 75% of 11 wins is 8.3 wins. fWAR has him at 8.8 WAR. So really, WAR overestimates the true effect that Guidry had on the Yankees winning record in 1978.

WAR is actually very accurate with great pitchers, which makes sense. It's a good blunt tool, so it can do a good job with the extremes. The difficulty WAR has, imo, is in the middle. It just has a hard time accurately telling who's been better over the last few years, Mike Leake or Homer Bailey or Matt Garza or Ryan Dempster or Bronson Arroyo.

[Brutus]

I don't have a problem with trying if the number is reliable. Pitching WAR is more reliable than a lot of stats that get thrown around all the time, but it also has its problems. For one, the impact of quality relievers is undervalued in relation to middle of the road starting pitchers.

Another issue is the reliance on FIP to accumulate WAR.

We know, because it's been shown time and again, that pitchers do have an ability to influence batted ball types and what happens in play -- to a degree. But WAR uses FIP, at least Fangraphs' version does, and that takes absolutely nothing into account in the field of play.

I don't think the problem has to do with boiling it down to a single number, as you said. It's good to be able to stack a player's value up against everyone in the game, whether at his position or at another position. However, it should be taken with a grain of salt until we have done a better job quantifying every player's worth. There's still some wiggle room with pitchers, and a lot of room with defense.

[jojo]

Another issue is the reliance on FIP to accumulate WAR.

We know, because it's been shown time and again, that pitchers do have an ability to influence batted ball types and what happens in play -- to a degree. But WAR uses FIP, at least Fangraphs' version does, and that takes absolutely nothing into account in the field of play.

I don't think the problem has to do with boiling it down to a single number, as you said. It's good to be able to stack a player's value up against everyone in the game, whether at his position or at another position. However, it should be taken with a grain of salt until we have done a better job quantifying every player's worth. There's still some wiggle room with pitchers, and a lot of room with defense.

Yet taking "it" into account would be pecking at the periphery. It just isn't a major criticism no matter how many times it is brought up.

[757690]

Yet taking "it" into account would be pecking at the periphery. It just isn't a major criticism no matter how many times it is brought up.

Then why not include it? Especially if the goal is to be a a precise as possible.

[jojo]

Then why not include it? Especially if the goal is to be a a precise as possible.

Why are you assuming including it would be more precise?

[757690]

Why are you assuming including it would be more precise?

We know that many pitchers produce different batted ball types. We know that different batter ball types lead to different run prevention. Ego, we know that including batted ball types will be more precise in telling us how pitchers prevent runs, than if we don't.

[jojo]

We know that many pitchers produce different batted ball types. We know that different batter ball types lead to different run prevention. Ego, we know that including batted ball types will be more precise in telling us how pitchers prevent runs, than if we don't.

Im confused. Randomly pick 40 or 50 pitchers and illustrate for us please.

[BluegrassRedleg]

325 innings and a 3.74 K/BB goes a very long way.

I can't imagine what guys felt like at the end of a season where they threw 325 innings.